B200 M5 stuck at 'Discover BIOS POST completion'

[datbui001]

I was upgrading ESXi on a series of B200 M5 blades (running 4.1.2a BIOS/FW version) to ESXi 7.0.3, the rest of the hosts didn't have a problem and were successful until this one. It's just stuck at 'Configuring and testing memory..' in the BIOS POST screen, resets itself and loops back to that screen.

In UCS it's stuck at task below:

Discover Bios Post Completion

In Progress

Waiting for BIOS POST completion from CIMC on server 4/3(FSM-STAGE:sam:dme:ComputeBladeDiscover:BiosPostCompletion)

32

50

Tried a re-ack, reset CMOS, reset CIMC, reset all memory errors and lastly moving it to another chassis-slot, but no luck. UCS Faults tab shows no hardware issues, only bios-post-completion and power-deploy-wait failed. 

[Steven Tardy]

I'm assuming this was NOT a firmware upgrade and only a ESXi upgrade. Right?
If it was a firmware upgrade, then trying the old/prior firmware might allow the server to boot.

There have been a few CDETS which resemble this issue, but most of the ones I can find are on older firmware (4.1.1c and 4.0.4l) but not 4.1.2a.

Are there any numbers/digits on the bottom right of the POST screen? If so, what are those numbers?

Without logs to review I might would try:

• "1 DIMM per CPU" to see if this is really a failed DIMM which isn't reporting that the DIMM is in fact failed.
• If doesn't work with "1 DIMM per CPU", then try another two DIMMs.

Beyond that I would need to review the logs to know what is happening under the hood.

If you have a support contract, then open a TAC case.